On 1/9/15 3:33 PM, Linda Walsh wrote:
> I have a little calculator interface that I use readline for.
> Imagine my surprise when I type 1+<TAB> and get a list of files
> to add to 1. Um... not ideal?
OK. If you don't want completion, bind TAB to self-insert.
>
> I add a time/date stamp to each line typed in.
> (HISTTIMEFORMAT="%m%d@%H%M%S:")
>
> In bash when I scroll back, I don't the time
> entries, but I do in my calculator. How can I filter
> them out?
Use read_history() or read_history_range() to read the history file. If
the perl code is reading lines from the history file itself, stop doing
that.
> Note, I do set some options, but seem to need to write the timestamp
> into the history file myself.
> $ra->{history_comment_char}='#';
> $ra->{history_expansion_char}="";
> $ra->{history_write_timestamps}=1;
> $ra->{history_quotes_inhibit_expansion}=1;
>
> histfile update writes the time itself to get the time in the .hist file...
>
> if ($histfile_h) {
> printf($histfile_h "#%d\n%s\n", CORE::time,$_) unless $_ eq 'quit';
> $histfile_h->flush;
> }
>
> i.e. the write_timestamps option seems to be having no effect -- unless
> it's writing another history file someplace else...?
If the perl code is using write_history() or append_history() and has
correctly reflected the history_write_timestamps variable, you don't need
to do anything else. If the `flush' method isn't using those functions,
then it needs to be updated.
--
``The lyf so short, the craft so long to lerne.'' - Chaucer
``Ars longa, vita brevis'' - Hippocrates
Chet Ramey, ITS, CWRU c...@case.edu http://cnswww.cns.cwru.edu/~chet/
